Senior AI / Full-Stack Engineer (LLM & SaaS Integration) ## Overview We are looking for a **high-calibre Senior AI / Full-Stack Engineer** to lead the development of an AI-powered assistant embedded within a multi-tenant SaaS procurement platform. This is not a typical “chatbot” role. You will design and build a **production-grade AI copilot** that integrates with real business workflows, handles sensitive financial data, and operates within a secure, multi-tenant environment. You will work closely with product leadership and an existing backend team to deliver a robust, scalable, and secure AI layer. --- ## What You’ll Be Building * An AI-powered assistant embedded into a SaaS platform * Natural-language interface for procurement analytics and insights * Secure orchestration layer connecting LLMs to internal APIs (no direct database access) * Structured tool-calling system for controlled data retrieval and actions * Chat UI with tables, summaries, and visual outputs * Auditability, guardrails, and enterprise-grade security patterns --- ## Key Responsibilities ### AI & LLM Engineering * Design and implement LLM-based workflows using OpenAI / Azure OpenAI * Build structured tool/function calling systems (no free-form execution) * Engineer prompts, schemas, and response formats for reliability and consistency * Handle hallucination mitigation, guardrails, and output validation * Optimize performance, latency, and cost ### Backend & Integration * Build an AI orchestration service (Python or Node.js) * Integrate with existing backend APIs (PHP-based platform) * Work with internal teams to define and enforce secure API contracts * Implement authentication context handling (user, tenant, permissions) * Ensure strict separation of concerns between AI layer and data layer ### System Design & Security * Design for multi-tenant SaaS environments with strict data isolation * Ensure no cross-tenant data leakage under any scenario * Implement audit logging and traceability for all AI interactions * Follow best practices for secure AI systems (prompt injection mitigation, least privilege access, etc.) ### Frontend (optional but valuable) * Build or support development of chat-based UI components * Render structured outputs (tables, charts, summaries) * Collaborate with frontend developers for seamless embedding --- ## Required Experience * 5–10+ years of software engineering experience * Strong backend development experience (Python, Node.js, or similar) * Hands-on experience with LLM APIs (OpenAI, Azure OpenAI, Anthropic, etc.) * Experience with tool/function calling and structured outputs * Strong understanding of API design and integration patterns * Experience working in SaaS environments (multi-tenant systems preferred) * Solid understanding of authentication, authorization, and RBAC * Experience designing production systems (not just prototypes) --- ## Strong Plus * Experience with procurement, finance, or enterprise SaaS systems * Experience with prompt evaluation, testing frameworks, or LLM QA * Familiarity with frameworks like LangChain, LangGraph, or Semantic Kernel (used in a controlled manner) * Experience with data visualization (charts, dashboards) * Experience working with legacy or PHP-based systems --- ## What We’re Looking For * A **builder** who can own end-to-end delivery * Someone comfortable working with ambiguity and shaping solutions * Strong problem-solver with attention to detail * Security-aware mindset (critical for this role) * Ability to work independently and collaborate with distributed teams --- ## Project Scope * Initial MVP focused on read-only analytics and insights * Integration with existing backend APIs (no direct DB access) * 6–10 week delivery timeline for MVP * Potential for long-term extension into advanced features (alerts, recommendations, workflow automation) --- ## Engagement * Contract or project-based (with potential extension) * Remote-friendly * Immediate start preferred --- ## How to Apply Please include: * Examples of LLM/AI systems you have built (production preferred) * Description of your role and architecture decisions * GitHub or portfolio (if available) * Brief note on how you would approach building a secure AI assistant in a multi-tenant SaaS